| Feb 3, 2026 |
Activation Oracles |
Federico |
| Feb 10, 2026 |
Weird generalizations |
Lukas |
| Feb 16, 2026 |
The Dead Salmons of AI Interpretability |
Andor |
| Mar 2, 2026 |
Narrow Finetuning Leaves Clearly Readable Traces in Activation Differences |
Annemette |
| Mar 9, 2026 |
Linear Representations can change over the course of a conversation |
Federico |
| Mar 23, 2026 |
EasySteer |
Gianluca |
| Mar 30, 2026 |
Thought Branches |
Andrea |
| Apr 13, 2026 |
Evaluating and Understanding Scheming Propensity in LLM Agents |
Filippo |
| Apr 20, 2026, 13:00 |
Steering Evaluation-Aware Language Models to Act Like They Are Deployed |
Lukas |
| Apr 27, 2026 |
Emergently Misaligned Language Models Show Behavioral Self-Awareness That Shifts With Subsequent Realignment |
Laurene Vaugrante (University of Stuttgart) |
| May 4, 2026 |
Cross-Architecture Model Diffing with Crosscoders: Unsupervised Discovery of Differences Between LLMs |
Federico |
| May 11, 2026 |
Screening new literature: Natural Language Autoencoders, Neural Geometry, and Manifold Steering |
* |
| June 1, 2026, 15:00 |
Main topic: Position: Model Collapse Does Not Mean What You Think + Side reading: Subliminal learning |
Anton |
| June 8, 2026 |
In-Training Defense against Emergent Misalignment in Language Models |
David Kaczér and Florian Mai (Lamarr Institute for ML/AI) |
| June 15, 2026 |
Subliminal Learning Is Steering Vector Distillation |
Lukas |
| June 22, 2026 |
NeuroFaith: Evaluating LLM Self-Explanation Faithfulness via Internal Representation Alignment |
Gianluca |
| June 29, 2026 |
Manifold Steering |
Johannes/Lukas |
| July 6, 2026 |
Solipsistic Superintelligence is Unlikely to be Cooperative |
Filippo |
| July 13, 2026 |
Human-AI Complementarity: A Goal for Amplified Oversight |
Shamim |
| August 3, 2026 |
The Jacobian Lens |
Annemette |
| August 24, 2026 |
A mechanistic analysis of looped reasoning models |
Andrea |
| Sept 14, 2026 |
A unifying framework from neural superposition to sparse interpretable codes |
Lukas |
| Sept 28, 2026 |
Closed-form predictive coding via hierarchical Gaussian filters |
Aleksandrs Baskakovs (Aarhus University) |
| Oct 26, 2026 |
Neural Networks Are Already Clean |
tbd |
| Nov 9, 2026 |
Weight Oracles: Reading Neural Network Weights with Language Models |
Federico |
| Nov 16, 2026 |
AuditBench: Evaluating Alignment Auditing Techniques on Models with Hidden Behaviors |
tbd |
| Nov 23, 2026 |
Training Honesty Adapters To Detect Deception In LLMs |
tbd |